Coinversaa Pulse
v0.5.0Crypto intelligence for AI agents. 7 free tools + 32 premium tools for Hyperliquid trader analytics, behavioral cohorts, syncer-backed risk data, live market...
⭐ 0· 416·0 current·0 all-time
byNassif Chedrawi@nazchedz
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Benign
medium confidencePurpose & Capability
The name/description (crypto intelligence, MCP-compatible tools) map to the actions in SKILL.md: it exposes MCP tools and documents adding an MCP server via npx @coinversaa/mcp-server and optionally providing COINVERSAA_API_KEY. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md stays on-topic: it documents free vs API-key tiers, lists tools and rate limits, and provides concrete instructions for adding an MCP server and setting COINVERSAA_API_KEY in agent configs. It does instruct editing local agent config files (e.g., Claude and Cursor config paths) and to run npx which will fetch and run remote package code — this is expected for this capability but is an action that executes third-party code and touches local app config files.
Install Mechanism
There is no packaged install spec, but the runtime instructions rely on npx -y @coinversaa/mcp-server. That will download and execute an npm package from the registry (supply-chain risk). This is proportionate for a connector but is a moderate operational risk compared to instruction-only skills that never fetch code.
Credentials
Only optional environment variables are documented (COINVERSAA_API_KEY, COINVERSAA_API_URL). COINVERSAA_API_KEY is optional (7 free tools without a key) and is appropriate for unlocking paid features. No unrelated tokens (AWS, SSH, wallets, etc.) are requested.
Persistence & Privilege
always:false and normal autonomous invocation are in effect. The skill suggests writing entries to local agent config files to register the MCP server — this is expected and limited to the user's agent configuration; the skill does not request system-wide privileges or to modify other skills.
Assessment
This skill appears internally consistent with its stated purpose, but review these points before installing:
- npx executes remote npm code: the SKILL.md instructs running `npx -y @coinversaa/mcp-server`. That will download and run code from the npm registry. If you are concerned about supply-chain risk, inspect the @coinversaa/mcp-server package source (repository listed as https://github.com/coinversaa/mcp-server), check maintainers, versions, and ideally review the package before running it.
- API key is optional: you can use 7 free tools without providing credentials. If you do provide COINVERSAA_API_KEY, only give it to this connector and avoid pasting keys into shared locations. The SKILL.md documents the key prefix (cvsa_) and an optional COINVERSAA_API_URL.
- Config file edits are local: the skill tells you how to add the MCP entry to local agent config files (Claude, Cursor, OpenClaw). These are expected setup steps but will change agent behavior by registering the connector — only perform if you trust coinversaa.ai and the npm package.
- Trading-related caution: the documentation mentions backend-signed orders and approvals on Hyperliquid. The skill does not request wallet keys, but if you plan agentic trading, understand who signs orders (their backend signer) and do not share private wallet keys or vault credentials unless you explicitly intend to enable trading with that service.
- If you want higher assurance: verify the npm package checksum/release on the repository, review the repository code, and prefer installing pinned versions rather than allowing npx to fetch the latest by default.
If you want, I can: summarize the SKILL.md fully, extract the exact list of tools and rate limits, or fetch the GitHub repo link to show where to audit the npm package.Like a lobster shell, security has layers — review code before you run it.
latestvk97f4bppmngwsxpwaehvcw358984bpjh
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
